preprocesspanel.java

Java 编写的多种数据挖掘算法 包括聚类、分类、预处理等

	      // write the data file
	      for (int i = 0; i < inst.numInstances(); i++) {
		Instance tempI = inst.instance(i);
		for (int j = 0; j < inst.numAttributes(); j++) {
		  if (j != classIndex) {
		    if (tempI.isMissing(j)) {
		      w2.write("?,");
		    } else if (inst.attribute(j).isNominal() || 
			       inst.attribute(j).isString()) {
		      w2.write(inst.attribute(j).value((int)tempI.value(j))+",");
		    } else {
		      w2.write(""+tempI.value(j)+",");
		    }
		  }
		}
		//		w2.write(inst.instance(i).toString());
		// write the class value
		if (tempI.isMissing(classIndex)) {
		  w2.write("?");
		} else {
		  w2.write(inst.attribute(classIndex).
			   value((int)tempI.value(classIndex)));
		}
		w2.write("\n");
	      }
	      w2.close();
	      m_Log.statusMessage("OK");
	 
	    } catch (Exception ex) {
	      ex.printStackTrace();
	      m_Log.logMessage(ex.getMessage());
	    }
	    m_IOThread = null;
	  }
	};
      m_IOThread.setPriority(Thread.MIN_PRIORITY); // UI has most priority
      m_IOThread.start();
    } else {
      JOptionPane.showMessageDialog(this,
				    "Can't save at this time,\n"
				    + "currently busy with other IO",
				    "Save c45 format",
				    JOptionPane.WARNING_MESSAGE);
    }
  }

  /**
   * Saves the current instances in binary serialized form to a file
   *
   * @param f a value of type 'File'
   * @param inst the instances to save
   */
  protected void saveSerializedInstancesToFile(final File f, 
					       final Instances inst) {
    if (m_IOThread == null) {
      m_IOThread = new Thread() {
	  public void run() {
	    try {
	      m_Log.statusMessage("Saving to file...");

	      ObjectOutputStream oos = 
		  new ObjectOutputStream(
		  new BufferedOutputStream(
		  new FileOutputStream(f)));

	      oos.writeObject(inst);
	      oos.flush();
	      oos.close();

	      m_Log.statusMessage("OK");
	    } catch (Exception ex) {
	      ex.printStackTrace();
	      m_Log.logMessage(ex.getMessage());
	    }
	    m_IOThread = null;
	  }
	};
      m_IOThread.setPriority(Thread.MIN_PRIORITY); // UI has most priority
      m_IOThread.start();
    } else {
      JOptionPane.showMessageDialog(this,
				    "Can't save at this time,\n"
				    + "currently busy with other IO",
				    "Save binary serialized instances",
				    JOptionPane.WARNING_MESSAGE);
    } 
  }

  /**
   * Saves the current instances to the supplied file.
   *
   * @param f a value of type 'File'
   * @param inst the instances to save
   * @param saveHeader true to save in arff format, false to save in csv
   */
  protected void saveInstancesToFile(final File f, final Instances inst,
				     final boolean saveHeader) {
      
    if (m_IOThread == null) {
      m_IOThread = new Thread() {
	public void run() {
	  try {
	    m_Log.statusMessage("Saving to file...");
	    Writer w = new BufferedWriter(new FileWriter(f));
	    if (saveHeader) {
	      Instances h = new Instances(inst, 0);
	      w.write(h.toString());
	      w.write("\n");
	    } else {
	      // csv - write attribute names as first row
	      for (int i = 0; i < inst.numAttributes(); i++) {
		w.write(inst.attribute(i).name());
		if (i < inst.numAttributes()-1) {
		  w.write(",");
		}
	      }
	      w.write("\n");
	    }
	    for (int i = 0; i < inst.numInstances(); i++) {
	      w.write(inst.instance(i).toString());
	      w.write("\n");
	    }
	    w.close();
	    m_Log.statusMessage("OK");
	  } catch (Exception ex) {
	    ex.printStackTrace();
	    m_Log.logMessage(ex.getMessage());
	  }
	  m_IOThread = null;
	}
      };
      m_IOThread.setPriority(Thread.MIN_PRIORITY); // UI has most priority
      m_IOThread.start();
    } else {
      JOptionPane.showMessageDialog(this,
				    "Can't save at this time,\n"
				    + "currently busy with other IO",
				    "Save Instances",
				    JOptionPane.WARNING_MESSAGE);
    }
  }

  /**
   * Pops up generic object editor with list of conversion filters
   *
   * @param f the File
   */
  private void converterQuery(final File f) {
    final GenericObjectEditor convEd = new GenericObjectEditor(true);

    try {
      convEd.setClassType(weka.core.converters.Loader.class);
      convEd.setValue(new weka.core.converters.CSVLoader());
      ((GenericObjectEditor.GOEPanel)convEd.getCustomEditor())
	.addOkListener(new ActionListener() {
	    public void actionPerformed(ActionEvent e) {
	      tryConverter((Loader)convEd.getValue(), f);
	    }
	  });
    } catch (Exception ex) {
    }

    PropertyDialog pd = new PropertyDialog(convEd, 100, 100);
  }

  /**
   * Applies the selected converter
   *
   * @param cnv the converter to apply to the input file
   * @param f the input file
   */
  private void tryConverter(final Loader cnv, final File f) {

    if (m_IOThread == null) {
      m_IOThread = new Thread() {
	  public void run() {
	    try {
	      cnv.setSource(f);
	      Instances inst = cnv.getDataSet();
	      setInstances(inst);
	    } catch (Exception ex) {
	      m_Log.statusMessage(cnv.getClass().getName()+" failed to load "
				 +f.getName());
	      JOptionPane.showMessageDialog(PreprocessPanel.this,
					    cnv.getClass().getName()+" failed to load '"
					    + f.getName() + "'.\n"
					    + "Reason:\n" + ex.getMessage(),
					    "Convert File",
					    JOptionPane.ERROR_MESSAGE);
	      m_IOThread = null;
	      converterQuery(f);
	    }
	    m_IOThread = null;
	  }
	};
      m_IOThread.setPriority(Thread.MIN_PRIORITY); // UI has most priority
      m_IOThread.start();
    }
  }

  /**
   * Loads results from a set of instances contained in the supplied
   * file. This is started in the IO thread, and a dialog is popped up
   * if there's a problem.
   *
   * @param f a value of type 'File'
   */
  public void setInstancesFromFile(final File f) {
      
    if (m_IOThread == null) {
      m_IOThread = new Thread() {
	public void run() {
	  String fileType = f.getName();
	  try {
	    m_Log.statusMessage("Reading from file...");
	    if (f.getName().toLowerCase().endsWith(Instances.FILE_EXTENSION)) {	    
	      fileType = "arff";
	      Reader r = new BufferedReader(new FileReader(f));
	      setInstances(new Instances(r));
	      r.close();
	    } else if (f.getName().toLowerCase().endsWith(CSVLoader.FILE_EXTENSION)) {
	      fileType = "csv";
	      CSVLoader cnv = new CSVLoader();
	      cnv.setSource(f);
	      Instances inst = cnv.getDataSet();
	      setInstances(inst);
	    } else if (f.getName().toLowerCase().endsWith(C45Loader.FILE_EXTENSION)) {
	      fileType = "C45 names";
	      C45Loader cnv = new C45Loader();
	      cnv.setSource(f);
	      Instances inst = cnv.getDataSet();
	      setInstances(inst);
	    } else if (f.getName().toLowerCase().
		       endsWith(Instances.SERIALIZED_OBJ_FILE_EXTENSION)
		       || f.getName().toLowerCase().endsWith(".tmp")) {
	      ObjectInputStream ois = 
		new ObjectInputStream(new BufferedInputStream(new FileInputStream(f)));
	      setInstances((Instances)ois.readObject());
	      ois.close();
	    } else {
	      throw new Exception("Unrecognized file type");
	    }
	  } catch (Exception ex) {
	    m_Log.statusMessage("File '" + f.getName() + "' not recognised as an "
				+fileType+" file.");
	    m_IOThread = null;
	    if (JOptionPane.showOptionDialog(PreprocessPanel.this,
					     "File '" + f.getName()
					     + "' not recognised as an "
					     +fileType+" file.\n"
					     + "Reason:\n" + ex.getMessage(),
					     "Load Instances",
					     0,
					     JOptionPane.ERROR_MESSAGE,
					     null,
					     new String[] {"OK", "Use Converter"},
					     null) == 1) {
	    
	      converterQuery(f);
	    }
	  }
	  m_IOThread = null;
	}
      };
      m_IOThread.setPriority(Thread.MIN_PRIORITY); // UI has most priority
      m_IOThread.start();
    } else {
      JOptionPane.showMessageDialog(this,
				    "Can't load at this time,\n"
				    + "currently busy with other IO",
				    "Load Instances",
				    JOptionPane.WARNING_MESSAGE);
    }
  }
  
  /**
   * Loads instances from a database
   *
   * @param iq the InstanceQuery object to load from (this is assumed
   * to have been already connected to a valid database).
   */
  public void setInstancesFromDB(final InstanceQuery iq) {
    if (m_IOThread == null) {
      m_IOThread = new Thread() {
	public void run() {
	  
	  try {
	    m_Log.statusMessage("Reading from database...");
	    final Instances i = iq.retrieveInstances();
	    SwingUtilities.invokeAndWait(new Runnable() {
	      public void run() {
		setInstances(new Instances(i));
	      }
	    });
	    iq.disconnectFromDatabase();
	  } catch (Exception ex) {
	    m_Log.statusMessage("Problem executing DB query "+m_SQLQ);
	    JOptionPane.showMessageDialog(PreprocessPanel.this,
					  "Couldn't read from database:\n"
					  + ex.getMessage(),
					  "Load Instances",
					  JOptionPane.ERROR_MESSAGE);
	  }

	   m_IOThread = null;
	}
      };

      m_IOThread.setPriority(Thread.MIN_PRIORITY); // UI has most priority
      m_IOThread.start();
    } else {
       JOptionPane.showMessageDialog(this,
				    "Can't load at this time,\n"
				    + "currently busy with other IO",
				    "Load Instances",
				    JOptionPane.WARNING_MESSAGE);
    }
  }

  /**
   * Loads instances from a URL.
   *
   * @param u the URL to load from.
   */
  public void setInstancesFromURL(final URL u) {

    if (m_IOThread == null) {
      m_IOThread = new Thread() {
	public void run() {

	  try {
	    m_Log.statusMessage("Reading from URL...");
	    Reader r = new BufferedReader(
		       new InputStreamReader(u.openStream()));
	    setInstances(new Instances(r));
	    r.close();
	  } catch (Exception ex) {
	    ex.printStackTrace();
	    m_Log.statusMessage("Problem reading " + u);
	    JOptionPane.showMessageDialog(PreprocessPanel.this,
					  "Couldn't read from URL:\n"
					  + u + "\n"
					  + ex.getMessage(),
					  "Load Instances",
					  JOptionPane.ERROR_MESSAGE);
	  }

	  m_IOThread = null;
	}
      };
      m_IOThread.setPriority(Thread.MIN_PRIORITY); // UI has most priority
      m_IOThread.start();
    } else {
      JOptionPane.showMessageDialog(this,
				    "Can't load at this time,\n"
				    + "currently busy with other IO",
				    "Load Instances",
				    JOptionPane.WARNING_MESSAGE);
    }
  }

  /**
   * Backs up the current state of the dataset, so the changes can be undone.
   */
  public void addUndoPoint() throws Exception {
    
    if (m_Instances != null) {
      // create temporary file
      File tempFile = File.createTempFile("weka", null);
      tempFile.deleteOnExit();

      ObjectOutputStream oos = 
	new ObjectOutputStream(
	new BufferedOutputStream(
	new FileOutputStream(tempFile)));
    
      oos.writeObject(m_Instances);
      oos.flush();
      oos.close();

      // update undo file list
      if (m_tempUndoFiles[m_tempUndoIndex] != null) {
	// remove undo points that are too old
	m_tempUndoFiles[m_tempUndoIndex].delete();
      }
      m_tempUndoFiles[m_tempUndoIndex] = tempFile;
      if (++m_tempUndoIndex >= m_tempUndoFiles.length) {
	// wrap pointer around
	m_tempUndoIndex = 0;
      }

      m_UndoBut.setEnabled(true);
    }
  }

  /**
   * Reverts to the last backed up version of the dataset.
   */
  public void undo() {

    if (--m_tempUndoIndex < 0) {
      // wrap pointer around
      m_tempUndoIndex = m_tempUndoFiles.length-1;
    }
    
    if (m_tempUndoFiles[m_tempUndoIndex] != null) {
      // load instances from the temporary file
      setInstancesFromFile(m_tempUndoFiles[m_tempUndoIndex]);

      // update undo file list
      m_tempUndoFiles[m_tempUndoIndex] = null;
    }
    
    // update undo button
    int temp = m_tempUndoIndex-1;
    if (temp < 0) {
      temp = m_tempUndoFiles.length-1;
    }
    m_UndoBut.setEnabled(m_tempUndoFiles[temp] != null);
  }
  
  /**
   * edits the current instances object in the viewer 
   */
  public void edit() {
    ViewerDialog        dialog;
    int                 result;
    Instances           copy;
    Instances           newInstances;
    
    final int classIndex = m_AttVisualizePanel.getColoringIndex();
    copy   = new Instances(m_Instances);
    copy.setClassIndex(classIndex);
    dialog = new ViewerDialog(null);
    result = dialog.showDialog(copy);
    if (result == ViewerDialog.APPROVE_OPTION) {
      try {
        addUndoPoint();
      }
      catch (Exception e) {
        e.printStackTrace();
      }
      // if class was not set before, reset it again after use of filter
      newInstances = dialog.getInstances();
      if (m_Instances.classIndex() < 0)
        newInstances.setClassIndex(-1);
      setInstances(newInstances);
    }
  }
  
  /**
   * Tests out the instance-preprocessing panel from the command line.
   *
   * @param args ignored
   */
  public static void main(String [] args) {

    try {
      final JFrame jf = new JFrame("Weka Explorer: Preprocess");
      jf.getContentPane().setLayout(new BorderLayout());
      final PreprocessPanel sp = new PreprocessPanel();
      jf.getContentPane().add(sp, BorderLayout.CENTER);
      weka.gui.LogPanel lp = new weka.gui.LogPanel();
      sp.setLog(lp);
      jf.getContentPane().add(lp, BorderLayout.SOUTH);
      jf.addWindowListener(new WindowAdapter() {
	public void windowClosing(WindowEvent e) {
	  jf.dispose();
	  System.exit(0);
	}
      });
      jf.pack();
      jf.setSize(800, 600);
      jf.setVisible(true);
    } catch (Exception ex) {
      ex.printStackTrace();
      System.err.println(ex.getMessage());
    }
  }
}